Advanced Lane Finding Project

The goals / steps of this project are the following:

  • Compute the camera calibration matrix and distortion coefficients given a set of chessboard images.
  • Apply a distortion correction to raw images.
  • Use color transforms, gradients, etc., to create a thresholded binary image.
  • Apply a perspective transform to rectify binary image ("birds-eye view").
  • Detect lane pixels and fit to find the lane boundary.
  • Determine the curvature of the lane and vehicle position with respect to center.
  • Warp the detected lane boundaries back onto the original image.
  • Output visual display of the lane boundaries and numerical estimation of lane curvature and vehicle position.

Camera Calibration

Camera Calibration using Chessboard images

In [1]:
import numpy as np
import cv2
import glob
import matplotlib.pyplot as plt
import pickle
import json

grid_x = 9
grid_y = 6
# prepare object points, like (0,0,0), (1,0,0), (2,0,0) ....,(6,5,0)
objp = np.zeros((grid_y*grid_x,3), np.float32)
objp[:,:2] = np.mgrid[0:grid_x, 0:grid_y].T.reshape(-1,2)

# Arrays to store object points and image points from all the images.
objpoints = [] # 3d points in real world space
imgpoints = [] # 2d points in image plane.
imgs = []

# Make a list of calibration images
img_fnames = glob.glob('camera_cal/calibration*.jpg')

# Step through the list and search for chessboard corners
for idx, fname in enumerate(img_fnames):
    img = cv2.imread(fname)
    gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)

    # Find the chessboard corners
    ret, corners = cv2.findChessboardCorners(gray, (grid_x, grid_y), None)

    # If found, add object points, image points
    if ret == True:
        objpoints.append(objp)
        imgpoints.append(corners)

        # Draw and display the corners
        cv2.drawChessboardCorners(img, (8,6), corners, ret)
        imgs.append(img)
        
h, w = imgs[0].shape[:2]
img_size = (w, h)
ret, mtx, dist, rvecs, tvecs = cv2.calibrateCamera(objpoints, imgpoints, img_size,None,None)

# Save the camera calibration result for later use (we won't worry about rvecs / tvecs)
np.savez('camera_cal.npz', mtx=mtx, dist=dist)

Visualization of Undistortion

In [2]:
%matplotlib inline

f, subplots = plt.subplots(2, 3, figsize=(20,10))

for idx, img in enumerate(imgs[:3]):
    dst = cv2.undistort(img, mtx, dist, None, mtx)
    img = cv2.cvtColor(img, cv2.COLOR_BGR2RGB)
    dst = cv2.cvtColor(dst, cv2.COLOR_BGR2RGB)
    subplots[0][idx].imshow(img)
    subplots[0][idx].set_title('Original Image', fontsize=10)
    subplots[1][idx].imshow(dst)
    subplots[1][idx].set_title('Undistorted Image', fontsize=10)

Thresholding

Pipeline on Test Images

In [3]:
def show_imgs(imgs1, imgs2, title1, title2):
    h = len(imgs1)
    f, subplots = plt.subplots(h, 2, figsize=(20, 5*h))

    for idx, img_pair in enumerate(zip(imgs1, imgs2)):
        img1, img2 = img_pair
        if len(img1.shape) == 3:
            img1 = cv2.cvtColor(img1, cv2.COLOR_BGR2RGB)
            subplots[idx][0].imshow(img1)
        else:
            subplots[idx][0].imshow(img1, cmap='gray')
        subplots[idx][0].set_title(title1, fontsize=20)
        if len(img2.shape) == 3:
            img2 = cv2.cvtColor(img2, cv2.COLOR_BGR2RGB)
            subplots[idx][1].imshow(img2)
        else:
            subplots[idx][1].imshow(img2, cmap='gray')
        subplots[idx][1].set_title(title2, fontsize=20)
In [4]:
import os

img_fnames = ['test_images/straight_lines1.jpg', 'test_images/straight_lines2.jpg',
             'test_images/test1.jpg', 'test_images/test2.jpg', 'test_images/test3.jpg',
             'test_images/test4.jpg', 'test_images/test4.jpg', 'test_images/test6.jpg',]
original_imgs = []
undistorted_imgs = []

for idx, fname in enumerate(img_fnames):
    img = cv2.imread(fname)
    original_imgs.append(img)
    
    dst = cv2.undistort(img, mtx, dist, None, mtx)
    undistorted_imgs.append(dst)
    out = os.path.join('./output_images/undistortion', os.path.basename(fname))
    cv2.imwrite(out, dst)

show_imgs(original_imgs[:3], undistorted_imgs[:3], "original", "undistorted")

Gradients Thresholding

X,Y Gradient

In [5]:
# Define a function that takes an image, gradient orientation,
# and threshold min / max values.
def abs_sobel_thresh(img, orient='x', sobel_kernel=3, thresh=(0, 255)):
    # Convert to grayscale
    gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
    # Apply x or y gradient with the OpenCV Sobel() function
    # and take the absolute value
    if orient == 'x':
        abs_sobel = np.absolute(cv2.Sobel(gray, cv2.CV_64F, 1, 0, ksize=sobel_kernel))
    if orient == 'y':
        abs_sobel = np.absolute(cv2.Sobel(gray, cv2.CV_64F, 0, 1, ksize=sobel_kernel))
    # Rescale back to 8 bit integer
    scaled_sobel = np.uint8(255*abs_sobel/np.max(abs_sobel))
    # Create a copy and apply the threshold
    binary_output = np.zeros_like(scaled_sobel)
    # Here I'm using inclusive (>=, <=) thresholds, but exclusive is ok too
    binary_output[(scaled_sobel >= thresh[0]) & (scaled_sobel <= thresh[1])] = 1

    # Return the result
    return binary_output
    
In [6]:
# Run the function
image = undistorted_imgs[0]
grad_binary = abs_sobel_thresh(image, orient='x', thresh=(20,100))
# Plot the result
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(24, 9))
f.tight_layout()
ax1.imshow(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))
ax1.set_title('Original Image', fontsize=20)
ax2.imshow(grad_binary, cmap='gray')
ax2.set_title('Thresholded Gradient', fontsize=20)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)

Magnitude of the Gradient

In [7]:
def mag_thresh(img, sobel_kernel=3, mag_thresh=(0, 255)):
    
    # Apply the following steps to img
    # 1) Convert to grayscale
    gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
    # 2) Take the gradient in x and y separately
    sobelx = cv2.Sobel(gray, cv2.CV_64F, 1, 0, ksize=sobel_kernel)
    sobely = cv2.Sobel(gray, cv2.CV_64F, 0, 1, ksize=sobel_kernel)
    # 3) Calculate the magnitude 
    abs_sobelxy = np.sqrt(np.square(sobelx) + np.square(sobely))
    # 4) Scale to 8-bit (0 - 255) and convert to type = np.uint8
    scaled_sobelxy = np.uint8(255*abs_sobelxy/np.max(abs_sobelxy))
    # 5) Create a binary mask where mag thresholds are met
    binary_output = np.zeros_like(scaled_sobelxy)
    binary_output[(scaled_sobelxy >= mag_thresh[0]) & (scaled_sobelxy <= mag_thresh[1])] = 1
    # 6) Return this mask as your binary_output image
    return binary_output
In [8]:
# Run the function
mag_binary = mag_thresh(image, sobel_kernel=3, mag_thresh=(30, 100))
# Plot the result
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(24, 9))
f.tight_layout()
ax1.imshow(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))
ax1.set_title('Original Image', fontsize=50)
ax2.imshow(mag_binary, cmap='gray')
ax2.set_title('Thresholded Magnitude', fontsize=50)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)

Direction of the Gradient

In [9]:
# Define a function to threshold an image for a given range and Sobel kernel
def dir_threshold(img, sobel_kernel=3, thresh=(0, np.pi/2)):
    # Grayscale
    gray = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY)
    # Calculate the x and y gradients
    sobelx = cv2.Sobel(gray, cv2.CV_64F, 1, 0, ksize=sobel_kernel)
    sobely = cv2.Sobel(gray, cv2.CV_64F, 0, 1, ksize=sobel_kernel)
    # Take the absolute value of the gradient direction, 
    # apply a threshold, and create a binary image result
    absgraddir = np.arctan2(np.absolute(sobely), np.absolute(sobelx))
    binary_output =  np.zeros_like(absgraddir)
    binary_output[(absgraddir >= thresh[0]) & (absgraddir <= thresh[1])] = 1

    # Return the binary image
    return binary_output
In [10]:
# Run the function
dir_binary = dir_threshold(image, sobel_kernel=15, thresh=(0.7, 1.1))
# Plot the result
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(24, 9))
f.tight_layout()
ax1.imshow(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))
ax1.set_title('Original Image', fontsize=50)
ax2.imshow(dir_binary, cmap='gray')
ax2.set_title('Thresholded Grad. Dir.', fontsize=50)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)

Combining Gradient-Thresholding

In [11]:
def gradient_threshold(image):
    ksize = 3
    # Apply each of the thresholding functions
    gradx = abs_sobel_thresh(image, orient='x', sobel_kernel=ksize, thresh=(20, 100))
    grady = abs_sobel_thresh(image, orient='y', sobel_kernel=ksize, thresh=(20, 100))
    mag_binary = mag_thresh(image, sobel_kernel=ksize, mag_thresh=(30, 100))
    dir_binary = dir_threshold(image, sobel_kernel=ksize, thresh=(0.7, 1.3))

    combined = np.zeros_like(dir_binary)
    combined[((gradx == 1) & (grady == 1)) | ((mag_binary == 1) & (dir_binary == 1))] = 1
    #combined[((gradx == 1) & (grady == 1))] = 1
    
    return combined

gradient_imgs = []
for idx, fname in enumerate(img_fnames):
    img = undistorted_imgs[idx]
    dst = gradient_threshold(img)
    gradient_imgs.append(dst)
    
    out = os.path.join('./output_images/gradient', os.path.basename(fname))
    cv2.imwrite(out, dst)
    
show_imgs(undistorted_imgs[:3], gradient_imgs[:3], "undistorted", "combined gradient thresholding")

HLS and Color Thresholds

In [12]:
# Define a function that thresholds the S-channel of HLS
def hls_select(img, thresh=(0, 255)):
    hls = cv2.cvtColor(img, cv2.COLOR_RGB2HLS)
    s_channel = hls[:,:,2]
    binary_output = np.zeros_like(s_channel)
    binary_output[(s_channel > thresh[0]) & (s_channel <= thresh[1])] = 1
    return binary_output
In [13]:
hls_binary = hls_select(image, thresh=(155, 255))

# Plot the result
f, (ax1, ax2) = plt.subplots(1, 2, figsize=(24, 9))
f.tight_layout()
ax1.imshow(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))
ax1.set_title('Original Image', fontsize=50)
ax2.imshow(hls_binary, cmap='gray')
ax2.set_title('Thresholded S', fontsize=50)
plt.subplots_adjust(left=0., right=1, top=0.9, bottom=0.)
In [14]:
hls_imgs = []
for idx, fname in enumerate(img_fnames):
    img = undistorted_imgs[idx]
    dst = hls_select(img, thresh=(150, 255))
    hls_imgs.append(dst)
    
    out = os.path.join('./output_images/hls', os.path.basename(fname))
    cv2.imwrite(out, dst)
    
show_imgs(undistorted_imgs[2:], hls_imgs[2:], "undistored", "hls thresholding")

Combining Color and Gradient

In [15]:
def combined_threshold(image):
    ksize = 3
    # Apply each of the thresholding functions
    gradx = abs_sobel_thresh(image, orient='x', sobel_kernel=ksize, thresh=(20, 100))
    grady = abs_sobel_thresh(image, orient='y', sobel_kernel=ksize, thresh=(20, 100))
    mag_binary = mag_thresh(image, sobel_kernel=ksize, mag_thresh=(30, 100))
    dir_binary = dir_threshold(image, sobel_kernel=15, thresh=(0.7, 1.3))
    
    hls_binary = hls_select(image, thresh=(130, 255))

    combined = np.zeros_like(dir_binary)
    combined[((gradx == 1) & (grady == 1)) | ((mag_binary == 1) & (dir_binary == 1)) | (hls_binary == 1)] = 1

    return combined

combined_imgs = []
for idx, fname in enumerate(img_fnames):
    img = undistorted_imgs[idx]
    dst = combined_threshold(img)
    combined_imgs.append(dst)
    
    out = os.path.join('./output_images/combined', os.path.basename(fname))
    cv2.imwrite(out, dst)
    
show_imgs(undistorted_imgs[2:], combined_imgs[2:], "undistored", "combined final")

Transform

In [16]:
W, H = 1500, 1280
trim_w, trim_h = 1500, 1000
src = np.array([[594, 450], [684, 450], [1056, 690], [250,690]], np.float32)
dst = np.array([[250, 0], [1056, 0], [1056, H], [250, H]], np.float32)

M = cv2.getPerspectiveTransform(src, dst)
Minv = cv2.getPerspectiveTransform(dst, src)

# Save the result for later use
np.savez('warp.npz', M=M, Minv=Minv)
In [17]:
def unwarp_trim(img):
    warped = cv2.warpPerspective(img, M, (W, H), flags=cv2.INTER_LINEAR)
    #delete the next two lines
    return warped[H-trim_h:]

def recover(img):
    h, w = img.shape[:2]
    new = np.zeros((H, W, 3), np.uint8)
    new[H-h:] = img
    
    return new

warped_imgs = []
for idx, fname in enumerate(img_fnames):
    img = undistorted_imgs[idx]
    dst = unwarp_trim(img)
    warped_imgs.append(dst)
    
    out = os.path.join('./output_images/warped', os.path.basename(fname))
    cv2.imwrite(out, dst)
    
show_imgs(undistorted_imgs[2:], warped_imgs[2:], "undistored", "warped")
In [18]:
def noise_filter(image):
    kernel = np.ones((5,5),np.uint8)
    image = cv2.erode(image,kernel,iterations = 1)
    image = cv2.dilate(image,kernel,iterations = 1)
    return image

binary_warped_imgs = []
for idx, fname in enumerate(img_fnames):
    img = undistorted_imgs[idx]
    img = combined_threshold(img)
    dst = unwarp_trim(img)
    
    dst = noise_filter(dst)

    binary_warped_imgs.append(dst)
    
    out = os.path.join('./output_images/binary_warped', os.path.basename(fname))
    cv2.imwrite(out, dst)
    
show_imgs(combined_imgs[2:], binary_warped_imgs[2:], "binary", "binary_warped")

Implement Sliding Windows and Fit a Polynomial

In [19]:
import numpy as np
import cv2
import matplotlib.pyplot as plt

# Define a class to receive the characteristics of each line detection
class Line():
    def __init__(self, name):

        self.name = name
        # numbers of interations
        self.n = 5
        # was the line detected in the last iteration?
        self.detected = False
        # x values of the last n fits of the line
        self.recent_xfitted = []
        # the last n fits of the line
        self.recent_fits = []
        #average x values of the fitted line over the last n iterations
        self.bestx = None
        #polynomial coefficients averaged over the last n iterations
        self.best_fit = None
        #polynomial coefficients for the most recent fit
        self.current_fit = None
        #radius of curvature of the line in some units
        self.radius_of_curvature = None
        #distance in meters of vehicle center from the line
        self.line_base_pos = None
        #difference in fit coefficients between last and new fits
        self.diffs = np.array([0,0,0], dtype='float')
        #x values for detected line pixels
        self.allx = None
        #y values for detected line pixels
        self.ally = None
        # threshold for difference
        self.threshold_fit = 700
        self.threshold_fitx = 200
        self.filter = ''

    def check_outlier(self, fit, fitx):

        self.diffs = np.abs(self.current_fit - fit)
        self.diffx = np.mean(np.abs(self.bestx - fitx))
        if self.name == self.filter:
            print(self.name)
            print(self.diffs, np.linalg.norm(self.diffs))
            print(self.diffx)

        if np.linalg.norm(self.diffs) > self.threshold_fit:
            return False
        if self.diffx > self.threshold_fitx:
            return False

        return True

    def set(self, fit, fitx):
        self.detected = True
        self.current_fit = fit
        self.recent_fits.append(fit)
        if len(self.recent_fits) > self.n:
            self.recent_fits.pop(0)
        self.best_fit = np.mean(np.vstack(self.recent_fits), axis=0)

        self.recent_xfitted.append(fitx)
        if len(self.recent_xfitted) > self.n:
            self.recent_xfitted.pop(0)

        self.bestx = np.mean(np.array(self.recent_xfitted), axis=0)

    def update(self, fit, ploty):
        y_eval = np.max(ploty)
        self.radius_of_curvature = ((1 + (2*fit[0]*y_eval + fit[1])**2)**1.5) / np.absolute(2*fit[0])
        fitx = fit[0]*ploty**2 + fit[1]*ploty + fit[2]
        if self.current_fit is not None:

            if self.check_outlier(fit, fitx):
                self.set(fit, fitx)
            else:
                if self.name == self.filter:
                    print("skip this frame")
                self.detected = False
                if len(self.recent_fits) > self.n:
                    self.recent_fits.pop(0)
        else:
            self.current_fit = fit
            self.set(fit, fitx)

        return self.bestx

class Lane():
    def __init__(self):
        self.left_fit = None
        self.right_fit = None
        self.MARGIN = 50
        self.size_threshold = 200

        self.left_line = Line("LEFT")
        self.right_line = Line("RIGHT")

    def find_lane_lines_first(self, binary_warped):
        # Assuming you have created a warped binary image called "binary_warped"
        # Take a histogram of the bottom half of the image
        histogram = np.sum(binary_warped[binary_warped.shape[0]//2:,:], axis=0)
        # Create an output image to draw on and  visualize the result
        out_img = np.dstack((binary_warped, binary_warped, binary_warped))*255
        # Find the peak of the left and right halves of the histogram
        # These will be the starting point for the left and right lines
        midpoint = np.int(histogram.shape[0]/2)
        leftx_base = np.argmax(histogram[:midpoint])
        rightx_base = np.argmax(histogram[midpoint:]) + midpoint

        # Choose the number of sliding windows
        nwindows = 9
        # Set height of windows
        window_height = np.int(binary_warped.shape[0]/nwindows)
        # Identify the x and y positions of all nonzero pixels in the image
        nonzero = binary_warped.nonzero()
        nonzeroy = np.array(nonzero[0])
        nonzerox = np.array(nonzero[1])
        # Current positions to be updated for each window
        leftx_current = leftx_base
        rightx_current = rightx_base
        # Set the width of the windows +/- margin
        margin = self.MARGIN
        # Set minimum number of pixels found to recenter window
        minpix = 50
        # Create empty lists to receive left and right lane pixel indices
        left_lane_inds = []
        right_lane_inds = []

        # Step through the windows one by one
        for window in range(nwindows):
            # Identify window boundaries in x and y (and right and left)
            win_y_low = binary_warped.shape[0] - (window+1)*window_height
            win_y_high = binary_warped.shape[0] - window*window_height
            win_xleft_low = leftx_current - margin
            win_xleft_high = leftx_current + margin
            win_xright_low = rightx_current - margin
            win_xright_high = rightx_current + margin
            # Draw the windows on the visualization image
            cv2.rectangle(out_img,(win_xleft_low,win_y_low),(win_xleft_high,win_y_high),(0,255,0), 2)
            cv2.rectangle(out_img,(win_xright_low,win_y_low),(win_xright_high,win_y_high),(0,255,0), 2)
            # Identify the nonzero pixels in x and y within the window
            good_left_inds = ((nonzeroy >= win_y_low) & (nonzeroy < win_y_high) & (nonzerox >= win_xleft_low) & (nonzerox < win_xleft_high)).nonzero()[0]
            good_right_inds = ((nonzeroy >= win_y_low) & (nonzeroy < win_y_high) & (nonzerox >= win_xright_low) & (nonzerox < win_xright_high)).nonzero()[0]
            # Append these indices to the lists
            left_lane_inds.append(good_left_inds)
            right_lane_inds.append(good_right_inds)
            # If you found > minpix pixels, recenter next window on their mean position
            if len(good_left_inds) > minpix:
                leftx_current = np.int(np.mean(nonzerox[good_left_inds]))
            if len(good_right_inds) > minpix:
                rightx_current = np.int(np.mean(nonzerox[good_right_inds]))

        # Concatenate the arrays of indices
        left_lane_inds = np.concatenate(left_lane_inds)
        right_lane_inds = np.concatenate(right_lane_inds)

        # Extract left and right line pixel positions
        leftx = nonzerox[left_lane_inds]
        lefty = nonzeroy[left_lane_inds]
        rightx = nonzerox[right_lane_inds]
        righty = nonzeroy[right_lane_inds]

        # Fit a second order polynomial to each
        left_fit = None
        right_fit = None
        if len(leftx) > self.size_threshold and len(lefty) > self.size_threshold:
            left_fit = np.polyfit(lefty, leftx, 2)
        if len(rightx) > self.size_threshold and len(righty) > self.size_threshold:
            right_fit = np.polyfit(righty, rightx, 2)

        return left_fit, right_fit

    def find_lane_lines_after(self, binary_warped, left_fit, right_fit):
        # Assume you now have a new warped binary image
        # from the next frame of video (also called "binary_warped")
        # It's now much easier to find line pixels!
        nonzero = binary_warped.nonzero()
        nonzeroy = np.array(nonzero[0])
        nonzerox = np.array(nonzero[1])
        margin = self.MARGIN
        left_lane_inds = ((nonzerox > (left_fit[0]*(nonzeroy**2) + left_fit[1]*nonzeroy + left_fit[2] - margin)) & (nonzerox < (left_fit[0]*(nonzeroy**2) + left_fit[1]*nonzeroy + left_fit[2] + margin)))
        right_lane_inds = ((nonzerox > (right_fit[0]*(nonzeroy**2) + right_fit[1]*nonzeroy + right_fit[2] - margin)) & (nonzerox < (right_fit[0]*(nonzeroy**2) + right_fit[1]*nonzeroy + right_fit[2] + margin)))

        # Again, extract left and right line pixel positions
        leftx = nonzerox[left_lane_inds]
        lefty = nonzeroy[left_lane_inds]
        rightx = nonzerox[right_lane_inds]
        righty = nonzeroy[right_lane_inds]
        # Fit a second order polynomial to each
        left_fit = None
        right_fit = None
        if len(leftx) > self.size_threshold and len(lefty) > self.size_threshold:
            left_fit = np.polyfit(lefty, leftx, 2)
        if len(rightx) > self.size_threshold and len(righty) > self.size_threshold:
            right_fit = np.polyfit(righty, rightx, 2)

        return left_fit, right_fit

    def find_lane_lines(self, binary_warped):
        if (self.left_fit is None) and (self.right_fit is None):
            left_fit, right_fit = self.find_lane_lines_first(binary_warped)
            if left_fit is not None:
                self.left_fit = left_fit
            if right_fit is not None:
                self.right_fit = right_fit
        else:
            #left_fit, right_fit = self.find_lane_lines_after(binary_warped, self.left_fit, self.right_fit)
            left_fit, right_fit = self.find_lane_lines_first(binary_warped)
            if left_fit is not None:
                self.left_fit = left_fit
            if right_fit is not None:
                self.right_fit = right_fit

        h, w = binary_warped.shape[:2]

        ploty = np.linspace(0, h-1, num=h) # to cover same y-range as image


#        left_fitx = self.left_fit[0]*ploty**2 + self.left_fit[1]*ploty + self.left_fit[2]
#        right_fitx = self.right_fit[0]*ploty**2 + self.right_fit[1]*ploty + self.right_fit[2]

        left_fitx = self.left_line.update(self.left_fit, ploty)
        right_fitx = self.right_line.update(self.right_fit, ploty)

        # Create an image to draw the lines on
        warp_zero = np.zeros_like(binary_warped).astype(np.uint8)
        color_warp = np.dstack((warp_zero, warp_zero, warp_zero))
        if left_fitx is not None and right_fitx is not None:

            # Recast the x and y points into usable format for cv2.fillPoly()
            pts_left = np.array([np.transpose(np.vstack([left_fitx, ploty]))])
            pts_right = np.array([np.flipud(np.transpose(np.vstack([right_fitx, ploty])))])
            pts = np.hstack((pts_left, pts_right))

            # Draw the lane onto the warped blank image
            cv2.fillPoly(color_warp, np.int_([pts]), (0,255, 0))

        return color_warp
In [35]:
line_found_imgs = []
for idx, fname in enumerate(img_fnames):
    img = binary_warped_imgs[idx]
    lane = Lane()
    dst = lane.find_lane_lines(img)
    line_found_imgs.append(dst)
    
    out = os.path.join('./output_images/line_found', os.path.basename(fname))
    cv2.imwrite(out, dst)
    
show_imgs(binary_warped_imgs[2:], line_found_imgs[2:], "warped", "line_found")
In [22]:
def inv_warp(img, line_img):
    h, w = img.shape[:2]
    
    line_img = recover(line_img)

    inv_warped = cv2.warpPerspective(line_img, Minv, (w, h), flags=cv2.INTER_LINEAR)
    
    result = cv2.addWeighted(img, 1, inv_warped, 0.3, 0)
    return result


final_imgs = []
for idx, fname in enumerate(img_fnames):
    img = undistorted_imgs[idx]
    line_img = line_found_imgs[idx]
    dst = inv_warp(img, line_img)
    final_imgs.append(dst)
    
    out = os.path.join('./output_images/final', os.path.basename(fname))
    cv2.imwrite(out, dst)
    
show_imgs(undistorted_imgs[2:], final_imgs[2:], "undistorted", "final")
In [23]:
lane = None

def pipeline(img):
    undistorted = cv2.undistort(img, mtx, dist, None, mtx)
    combined = combined_threshold(undistorted)
    binary_warped = unwarp_trim(combined)
    line_img = lane.find_lane_lines(binary_warped)
    result = inv_warp(img, line_img)
    
    return result

Pipeline on Video

In [24]:
# Import everything needed to edit/save/watch video clips
from moviepy.editor import VideoFileClip
from IPython.display import HTML
In [25]:
def process_image(image):
    # NOTE: The output you return should be a color image (3 channel) for processing video below
    # TODO: put your pipeline here,
    # you should return the final output (image where lines are drawn on lanes)
    #result = lane_trace(image)
    result = pipeline(image)
    return result
In [26]:
lane = Lane()
white_output = 'result_project_video.mp4'
clip1 = VideoFileClip("project_video.mp4")
white_clip = clip1.fl_image(process_image) #NOTE: this function expects color images!!
%time white_clip.write_videofile(white_output, audio=False)
[MoviePy] >>>> Building video result_project_video.mp4
[MoviePy] Writing video result_project_video.mp4
100%|█████████▉| 1260/1261 [08:27<00:00,  2.43it/s]
[MoviePy] Done.
[MoviePy] >>>> Video ready: result_project_video.mp4 

CPU times: user 7min 58s, sys: 2min 10s, total: 10min 9s
Wall time: 8min 27s
In [27]:
HTML("""
<video width="960" height="540" controls>
  <source src="{0}">
</video>
""".format(white_output))
Out[27]:

Pipeline on Optional Videos

Challenge

In [33]:
img_fnames2 = ['test_images/challenge1.png', 'test_images/challenge2.png',
             'test_images/challenge3.png', 'test_images/challenge4.png']

def hls_select_v2(img, thresh=(0, 255)):
    hls = cv2.cvtColor(img, cv2.COLOR_RGB2HLS)
    s_channel = hls[:,:,2]
    l_channel = hls[:,:,1]
    kernel = np.ones((11,11),np.uint8)

    binary_output = np.zeros_like(s_channel)
    binary_output[(s_channel > thresh[0]) & (s_channel <= thresh[1]) | (l_channel >= 200)] = 1
    binary_output = cv2.dilate(binary_output,kernel,iterations = 1)
    return binary_output

def combined_threshold_v2(image):
    # Apply each of the thresholding functions
    gradx = abs_sobel_thresh(image, orient='x', sobel_kernel=5, thresh=(20, 100))
    grady = abs_sobel_thresh(image, orient='y', sobel_kernel=5, thresh=(20, 100))
    mag_binary = mag_thresh(image, sobel_kernel=5, mag_thresh=(30, 100))
    dir_binary = dir_threshold(image, sobel_kernel=15, thresh=(0.7, 1.3))

    hls_binary = hls_select_v2(image, thresh=(100, 255))

    combined = np.zeros_like(dir_binary)

    combined[(hls_binary == 1)&(((gradx == 1) & (grady == 1)) | ((mag_binary == 1) & (dir_binary == 1))) ] = 1

    return combined

original_imgs2=[]
processed_imgs = []

for idx, fname in enumerate(img_fnames2):
    lane = Lane()
    img = cv2.imread(fname)
    original_imgs2.append(img)
    
    dst = combined_threshold_v2(img)

    processed_imgs.append(dst)

show_imgs(original_imgs2, processed_imgs, "original", "undistorted")
In [34]:
lane = None

def pipeline(img):
    undistorted = cv2.undistort(img, mtx, dist, None, mtx)
    combined = combined_threshold(undistorted)
    binary_warped = unwarp_trim(combined)
    line_img = line.find_lane_lines(binary_warped)
    result = inv_warp(img, line_img)
    
    return result

def pipeline_v2(img):
    undistorted = cv2.undistort(img, mtx, dist, None, mtx)
    combined = combined_threshold(undistorted)
    binary_warped = unwarp_trim(combined)
    line_img = lane.find_lane_lines(binary_warped)
    result = inv_warp(img, line_img)
    
    return result

def process_image(image):
    result = pipeline(image)
    return result
In [32]:
lane = Lane()
challenge_output = 'result_challenge_video.mp4'
#clip2 = VideoFileClip('challenge_video.mp4')
#challenge_clip = clip2.fl_image(process_image)
#%time challenge_clip.write_videofile(challenge_output, audio=False)
HTML("""
<video width="960" height="540" controls>
  <source src="{0}">
</video>
""".format(challenge_output))
Out[32]: